N-benzoyl amino acids as LFA-1/ICAM inhibitors 1: Amino acid structure-activity relationship

Article abstract of DOI:10.1016/S0960-894X(03)00084-2

The association of ICAM-1 with LFA-1 plays a critical role in several autoimmune diseases. N-2-Bromobenzoyl L-tryptophan, compound 1, was identified as an inhibitor to the formation of the LFA-1/ICAM complex. The SAR of the amino acid indicates that the carboxylic acid is required for inhibition and that L-histidine is the most favored amino acid.

Full text of DOI:10.1016/S0960-894X(03)00084-2

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1015–1018
N-Benzoyl Amino Acids as LFA-1/ICAM Inhibitors 1:
Amino Acid Structure–Activity Relationship
Daniel J. Burdick,a,* Ken Paris,a Kenneth Weese,a Mark Stanley,a Maureen Beresini,b
Kevin Clark,b Robert S. McDowell,a,y James C. Marsters, Jr.a and Thomas R. Gadeka
aDepartment of Bioorganic Chemistry, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, USA
bBioanalytical Research and Development, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, USA
Received 7 November 2002; revised 15 January 2003; accepted 22 January 2003
Abstract—The association of ICAM-1 with LFA-1 plays a critical role in several autoimmune diseases. N-2-Bromobenzoyl
l-tryptophan, compound 1, was identified as an inhibitor to the formation of the LFA-1/ICAM complex. The SAR of the amino
acid indicates that the carboxylic acid is required for inhibition and that l-histidine is the most favored amino acid.
# 2003 Elsevier Science Ltd. All rights reserved.
Leukocyte function-associated antigen-1 (LFA-1, aLb2,
CD11a/CD18) is a member of the b2 integrin family.1,2
Expressed exclusively on leukocytes, LFA-1 interacts
with intracellular adhesion molecules 1, 2, and 3
(ICAM’s -1, -2, -3) in a variety of cell adhesion events
required for normal and pathologic functions of the
immune system.3ꢀ5 The interaction of LFA-1 with
ICAM-1 has been implicated in numerous autoimmune
diseases such as psoriasis, asthma, rheumatoid arthritis
and graft rejection and it is clear that an antagonist to
this interaction would play a critical therapeutic role in
the treatment of these maladies.6ꢀ8 Several monoclonal
antibodies and small-molecule antagonists have been
reported to address this medical need.9ꢀ11
This communication will focus on the SAR of the
amino acid moiety concentrating on the carboxylic acid
and the amino acid side chain. A subsequent communi-
cation will address the SAR of the 2-bromobenzoyl
moiety.
Two simple series of molecules were envisioned (Fig. 2)
to evaluate the importance of the carboxy terminus.
Series 1 (2–5) was designed to determine if the car-
boxylic acid could be replaced or modified. Series 2
We have recently reported our efforts to generate potent
LFA-1/ICAM antagonists.12,13 Mutagenesis of the first
domain of ICAM-1 identified six residues, in particular
glutamic acid 34 and tyrosine 66, that are essential for
the interaction with LFA-114 and we have incorporated
these critical amino acids in cyclic peptide antagonists.
The screening hit N-2-bromobenzoyl l-tryptophan (1)
(Fig. 1) was of interest not only because it was reported
to competitively inhibit the binding of ICAM-1 to LFA-1
with an IC50 of 1.7 mM, but because it contained ele-
ments of the peptide structure–activity relationship
(SAR).12 This similarity to the protein and our cyclic
peptides led us to evaluate the SAR of compound 1.
Figure 1.
*Corresponding author. Tel.: +1-650-225-1368; fax: +1-650-225-
2061; e-mail: burdick.dan@gene.com
yCurrent address: Sunesis Pharmaceuticals, 341 Oyster Point Boule-
vard, South San Francisco, CA 94080, USA.
Figure 2.
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00084-2

1016
D. J. Burdick et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1015–1018
(6–42) was proposed to determine the effect of replacing
the l-tryptophan with various tryptophan analogues,
unnatural aromatic amino acids and naturally occurring
l- and d-amino acids (R).
results in a>580-fold loss in potency. The same effect is
seen when the carboxylic acid is completely removed 4.
Even though hydroxamic acids are carboxylic acid sur-
rogates, compound 5 showed only weak inhibition.
These results agree with the protein mutagenesis data
which showed that glutamic acid 34 was essential for
binding.
Compounds 3–13, 35–39, and 42 were synthesized by
the route shown in Scheme 1 except where noted in the
tables. Fischer esterification of the appropriate com-
mercially available amino acid was done in methanol
followed by an extractive workup. The amine was acy-
lated with o-bromobenzoic acid using EDC, HOBT,
and DIPEA in DMF. After purification by silica chro-
matography using 5% methanol in DCM as the eluent,
the ester was removed with LiOH in THF/water. The
crude product was purified by reverse phase HPLC,
lyophilized and its molecular weight verified by electro
spray mass spectroscopy.
Series 2 was proposed to determine the effect of mod-
ifying the tryptophan indole (Table 2), to evaluate the
influence of the amino acid stereo chemistry and deter-
mine the contribution of the amino acid side chain
(Table 3).
Table 1. ELISA assay results for compounds 1–5 of Series 1 (Fig. 2)
demonstrating the effect of modifying the carboxylate
Compd
Y
IC50 (mM)
Scheme
2
Compounds 1, 2, 14–34, and 40–41 were synthesized by
standard Fmoc solid-phase synthetic techniques
(Scheme 2) on the appropriate commercially available
Fmoc amino acid 4-hydroxymethylphenoxy resin
(Wang resin).15,16 Following removal of the Fmoc
group with 20% piperidine in DMF and rinsing of the
resin, o-bromobenzoic acid was coupled using HBTU,
HOBT, and DIPEA in DMF. After drying the resin to a
constant weight, it was treated with a solution of TFA
containing 2.5% water and 2.5% anisole. Excess TFA
was removed in vacuo and the cleaved molecule was
extracted fromthe resin with a 10% solution of acetic
acid in water. The crude product was purified by
reverse-phase HPLC, lyophilized and its molecular
weight verified by electro spray mass spectroscopy.
1
1.7
2
3
>1000
>1000
2a
1b
4
5
H
>1000
1c
140
1d
aFmoc Rink resin used instead of Wang resin.
bMethyl ester not cleaved.
cTryptamine coupled, then purified.
dHydroxylamine coupled after methyl ester saponification, then purified.
All compounds were evaluated in an ICAM-1/LFA-1
enzyme linked immunosorbant assay (ELISA). Potency
is calculated as the concentration of inhibitor needed to
inhibit 50% of the ICAM-1/LFA-1 complex formation
(IC50). Reported IC50s are the arithmetic average of at
least two assays.
Table 2. ELISA assay results for compounds 1 and 6–11 of Series 2
(Fig. 2) demonstrating the effect of substituents on the indole of tryp-
tophan
Compd
R
IC50 (mM)
Scheme
2
1
l-Trp
1.7
Series 1 compounds were tested to determine the
importance of the carboxylic acid of the inhibitors. It
can be clearly seen in the data of Table 1, that the car-
boxylic acid is essential for inhibition. Converting the
carboxylate to the primary amide 2 or to the ester 3
63.8
28.4
10.8
1
1
1
6
7
8
Scheme 1. Reagents and conditions: (a) HCl (g), MeOH, 0 ꢁC to rt, 18 h;
(b) 2-bromobenzoic acid, EDC, HOBT, DIPEA, DMF, 2 h, rt;
(c) LiOH, THF, water, 1 h, rt.
9
7.64
2.77
1
1
1
10
11
24.15
Scheme 2. Reagents and conditions: (a) 20% piperidine/DMF, rt;
(b) 2-bromobenzoic acid, HBTU, HOBT, DIPEA, DMF, 2 h, rt;
(c) triisopropylsilane, water, TFA, 1 h, rt.

D. J. Burdick et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1015–1018
1017
Table 3. ELISA assay results for compounds 1 and 12–42 of Series 2
(Fig. 2) demonstrating the effect of different natural and unnatural l-
and d-amino acids
l-glutamic acid (27) were not potent. Only when they
were changed to their respective amide forms (28 and
29) did their potency improve.
Compd
Amino acid
IC50 (mM)
Scheme
The aromatic amino acid side chains were more potent
as a group than the alkyl side chains and in some cases
more potent than l-tryptophan. l-Phenylalanine (34)
was 2.8 times less potent than l-tryptophan. Interest-
ingly, lengthening the l-phenylalanine side chain by one
methylene (35) did not change its potency. Extending
the aromaticity with l-1-napthylalanine (36) and
l-2-napthylalanine (37) decreased the potency compared
to l-tryptophan 9.3- and 14.2- fold respectively. para
Phenyl (38) and para cyano substitutions (39) resulted in a
29.4- and 21.4-fold decreases in potency. The para
hydroxy substitution of l-tyrosine (40), however, showed
a 1.6-fold increase in potency over l-tryptophan.
Replacing the indole of l-tryptophan with thiophene
(41) reduced the potency by 3.2-fold, but replacement
with imidazole (42) improved inhibition 2.3-fold.
1
l-Trp
d-1-Nal
d-2-Nal
d-Ala
1.72
>200
242.5
57.45
280
2
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
2
1
1
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
d-Phe
d-Ser
72.6
d-Trp
76.55
>1000
6.64
l-Pro
l-Ala
l-Ile
106
l-Leu
54.25
11.33
313.5
2.71
1.21
>200
30.1
l-Val
Gly
l-Arg
l-Lys
l-Asp
l-Glu
l-Asn
1.08
1.77
6.15
3.84
4.26
7.46
4.89
5.09
15.95
24.4
50.5
36.8
1.09
5.5
l-Gln
l-Ser
Based on these and previously reported results, we con-
clude that within the class of 2-bromobenzoyl amino
acid antagonists the carboxylic acid is required for
inhibition of the LFA-1/ICAM-1 complex.12 This result
follows the trend that was observed in the ICAM-1
mutagenesis, which showed that the carboxylic acid of
the glutamic acid 34 side chain is critical for binding.
The inhibitors prefer to have an l-amino acid to a
d-amino acid regardless of the amino acid side chain.
And finally, even though l-histidine proved to be the
best amino acid side chain, l-tryptophan, l-asparagine,
l-lysine and l-tyrosine were of comparable potency.
l-Thr
l-Cys
l-Met
l-Phe
l-Hfe
l-1-Nal
l-2-Nal
l-Bip
l-Phe(4-CN)
l-Tyr
l-Thienylalanine
l-His
0.75
As shown in Table 2, any substitution on the indole of
l-tryptophan (1) reduces the potency of the inhibitors.
Replacing the carbon at the 7-position with a nitrogen
(6) results in a 37-fold decrease in potency. Adding a
halogen at the 5- (7, 8) or 6- (9) position of the indole
reduces potency 16.7-, 6.3- and 4.4-fold, respectively. A
methoxy group at the 5-position (10) is 8.5 times more
potent than a hydroxy substitution (11), but it is still
2-fold less potent than the unsubstituted case.
Low molecular weight inhibitors of large protein/pro-
tein interactions are one of the aims of the pharma-
ceutical industry. We have demonstrated that simple
changes of amino acid side chains and stereochemistry
can play a large role in the ability of these low molecular
weight molecules to inhibit the formation of the ICAM-1/
LFA-1 complex. A subsequent paper will explore the
SAR of the 2-bromobenzoyl moiety.
Table 3 shows various natural and unnatural amino
acid changes. In all cases, it is clear that the d-amino
acids tested (12–17) are less favored than the l-enantio-
mers (1, 19, 30, 34, 36, and 37) regardless of the side
chain. This preference may be due, in part, to the fact
that the l-amino acids direct the carboxylic acid to a
preferred point of contact while the d-amino acids do
not. Of the l-amino acids, l-proline (18) was the poor-
est inhibitor. Possible explanations include: the pyrroli-
dine ring does not allow the carboxylic acid to interact
with its point of contact; or the lack of amide bond
hydrogen leads to lower affinity.
Acknowledgements
The authors would like to thank Martin Struble and his
group for purification of the compounds.
References and Notes
1. Hynes, R. O. Cell 1987, 48, 549.
2. Mazzone, A.; Ricevute, G. Haematology 1995, 80, 161.
3. Springer, T. A. Nature 1990, 346, 425.
4. Carlos, T. M.; Harlan, J. M. Blood 1994, 84, 2068.
5. Gahmberg, C. G.; Tolvanen, M.; Kotovuori, P. Eur. J.
Biochem. 1997, 245, 215.
6. Bevilacqua, M. P.; Nelson, R. M.; Mannori, G. Ann. Rev.
Med. 1994, 45, 361.
7. Cornejo, C. J.; Winn, R. K.; Harlan, J. M. Adv. Pharmacol.
1997, 39, 99.
Compared to l-tryptophan (1), the alkyl amino acids
were not as potent. l-Alanine (19) was 4-fold less potent
than l-tryptophan, but it was 1.7- to 16-fold more
potent than the branched amino acids (20–22). The
basic side chains of l-arginine (24) and l-lysine (25)
were of comparable potency to l-tryptophan. Con-
versely, the acidic side chains of l-aspartic acid (26) and

1018
D. J. Burdick et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1015–1018
8. Henricks, P. A.; Nijkamp, F. P. Eur. J. Pharmacol. 1998,
344, 1.
9. Liu, G. Drugs Future 2001, 26, 767.
10. Liu, G. Expert Opin. Ther. Pat. 2001, 11, 1383.
11. Yun, W.; Desai, B. B.; Du, Y.; Fotouhi, N.; Gillespie, P.;
Guthrie, R. W.; Huang, K.; Kolinsky, K.; Li, S. H.; Mennona,
F. A.; Perrota, A.; Pietranico-Cole, S. L.; Portland, L.; Ver-
meulen, J. Abstracts of Papers, 224 National Meeting of the
American Chemical Society, Boston, MA; American Chemical
Society: Washington, DC, 2002.
M. S.; Marsters, J. C., Jr.; Paris, K. J.; Oare, D. A.; Reynolds,
M. E.; Ladner, C.; Zioncheck, K. A.; Lee, W. P.; Gribling, P.;
Dennis, M. S.; Skelton, N. J.; Tumas, D. B.; Clark, K. R.;
Keating, S. M.; Beresini, M. H.; Tilley, J. W.; Presta, L. G.;
Bodary, S. C. Science 2002, 295, 1086.
13. Burdick, D. J. In PCT Int. Appl. WO9949856, 1999.
14. Fisher, K. L.; Lu, J.; Riddle, L.; Kim, K. J.; Presta, L. G.;
Bodary, S. C. Mol. Biol. Cell 1997, 8, 501.
15. Carpino, L.; Han, G. J. Org. Chem. 1981, 37, 3404.
16. Fields, G.; Noble, R. L. Int. J. Peptide Protein Res. 1990,
35, 161.
12. Gadek, T. R.; Burdick, D. J.; McDowell, R. S.; Stanley,